Operation, Administration, and Maintenance (OAM) describes processes, activities, tools, standards, etc., involved with operating, administering, and maintaining computer networks. OAM is implemented for many different types of computer networks, such Ethernet, Internet Protocol (IP), multi-protocol label switching (MPLS), asynchronous transfer mode (ATM), virtual private LAN service (VPLS), etc. Each type of network will have a different type of OAM. For example, in an Ethernet network, 802.1ag Connectivity Fault Management (CFM) is used as an OAM management tool, and in a MPLS/VPLS network Virtual Circuit Connectivity Verification (VCCV) is used as an OAM tool. OAM tools are used, in part, to detect network connectivity faults, herein referred to as faults. A fault can be when a network device loses connectivity to the network. OAM tools detect a fault state change, such when a device becomes active on the network, by connecting to the network or becoming active on the network. Furthermore, a fault state change can be when a network loses connectivity to the network.
FIG. 1 illustrates one embodiment of network 100 CFM across a switched Ethernet and VPLS/MPLS environment. Network 100 is a heterogeneous network comprising two switched Ethernet networks 116A-B interconnected by a VPLS service provider network 106. In FIG. 1, Ethernet networks 116A-B are geographically dispersed. VPLS is a way to provide Ethernet based multipoint-to-multipoint communication over IP/MPLS networks. It allows geographically dispersed sites, such as Ethernet networks 116A-B to share an Ethernet broadcast domain by connecting sites through psuedowires. In VPLS, the LAN of each Ethernet network 116A-B is extended to edge of service provider network 106 and service provider network 106 emulates a switch or bridge to connect the customer LANs of Ethernet networks 116A-B to create a single bridged LAN. Coupling switched Ethernet networks 116A-B and VPLS network 106 at the edge of each network are network elements 104A-B. While in one embodiment, network elements 104A-B are edge routers, in alternative embodiments, network elements 104A-B are the same and/or different type of network element (switch, router, core router, etc.) For example, network element 104A couples switched Ethernet network 116A and VPLS network 106, while network element 104B couples switched Ethernet network 116B and VPLS network 106.
Each of switched Ethernet networks comprises maintenance endpoints (MEPs) coupled to the edge network elements. An MEPs is an actively managed CFM entity which can generate & receive CFM messages and track any responses such as personal computers, servers, bridges, switches, and other possible devices participating in Ethernet network. As illustrated in FIG. 1, Ethernet network 116A comprises maintenance endpoints 102A-C coupled to network element 104A and Ethernet network 116B comprises maintenance endpoints 102D-F coupled to network element 104B.
VPLS network 106 comprises network elements 108A-D, where network elements 108A and C couple to network element 104A and network elements 108B and D couple to network element 104B. Network elements 108A-B forward traffic between network elements 104A-B with pseudowire 110A. Network elements 108C-D forward traffic between network elements 104A-B with pseudowire 110B.
CFM is a standard that specifies protocols and protocol entities within virtual LAN (VLAN) aware bridges (such as network elements 104A-B) that provides capabilities for detecting, verifying, and isolating faults in VLANs. These capabilities can be used in networks operated by multiple independent organizations, each with restricted management access to each other's equipment. CFM defines a maintenance domain that as a network or part of the network for which faults in connectivity can be managed. Within each maintenance domain, there are several MEPs. An MEP is an actively managed CFM entity, which initiates, terminates, and reacts to CFM flows associated within a specific maintenance domain. Periodically, each MEP sends connectivity check messages to the other MEP in the maintenance domain. The connectivity check message is a multicast, unidirectional heartbeat that signals that the MEP sending the connectivity check message is up and coupled to the network. An MEP sending an initial connectivity check message signifies to other MEPs in the maintenance domain that the MEP has become active in this domain. Lack of connectivity check message from a particular MEP indicates to the other MEPs that the particular MEP is down or not participating in the domain.
In FIG. 1, MEPs 102A-C and D-F periodically send out connectivity check messages 112A-C and 114A-C, respectively. Each MEP can send out connectivity check messages every 3.3 milliseconds. These messages are multicast to all the other MEPs in the maintenance domain. Because the MEPs of the different Ethernet networks 116A-B belong to the same maintenance domain, the connectivity check messages are transmitted across VPLS network 106 via network elements 104AB to the MEPs in different Ethernet networks 116AB. In addition, because the connectivity check messages are multicast, network elements 104AB broadcast these messages to each of the pseudowires coupled to the respective network element 104AB in VPLS network 106.
Although CFM can provide an end-to-end fault management scheme for heterogeneous network 100, CFM is not scalable because of the flooding of the connectivity check (CC) messages by network elements 104A-B. As the number of MEPs in the maintenance domain increase, the amount of CC messages transmitted across VPLS network increases dramatically.